The Tunneling Percolation Staircase€¦ · The Tunneling Percolation Staircase Isaac Balberg Racah...

The Tunneling Percolation Staircase

Isaac Balberg

Racah Institute of Physics,

The Hebrew University, Jerusalem, Israel

Stat. Mech. Day IV. 230611

GRANULAR METAL COMPOSITES

σ (x-xc)t

t = 1.9

Thermal switches

Airplane wheels

Self regulated

heaters

Papyrus glyphs

CB

CB polymer composites and applications

Two fundamental questions

1) What determines the conductivity threshold?

2) What determines the shape of the σ(Φ) dependence?

Phys. Rev. B, 30, 3933 (1984): 304 Q

Confirmation of universality

0.08 0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.2410

-4

10-3

10-2

10-1

100

(

cm

)-1

CB volume fraction

t=1.95 , xc=0.10

XC72 2011

t = 1.7

tun (D = 3) ≈ 2

t

cpp )(

Experimentally observed non universal exponents

(p-pc)t

t = 6.4

vc = 39 vol.%

p = v/vc

R (≥ R0L1)

RL= R(L/)/(L/)D-1

R0

(p-pc)-

<R> (p-pc)-(t-tun)

5 10 15 20 25 30 35 400

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

r (arb.units)

h(r

) or

g(r

)

g(r) = g0exp[-2r/ξ]

h(r) = (1/d)exp[-r/d]

f(g) g(ξ/2d-1)

The diverging normalized distribution

in the tunneling percolation problem

f(g) =h(r)(dr/dg)

0gpc←p

d

ξ

t = tun+(2d/ξ)-1Phys. Rev. Lett, 59, 1305 (1987): 215 Q

σ (p-pc)t

Various carbon blacks

t = 6.4 t = 4 t =1.8

10 wt. %39 vol. %

Percolation and Tunneling?

V V

g(rij) = g0 exp[-2(rij-D)/ξ] What is the meaning of vc?

Carbon, 40, 139 (2002): 100 Q.

The hopping model

(the CPA network)

rc

Bc=(4π/3)[rc3- (2b)3]N

The increase

of N

conserves

topology but

increase the

value of g

Local

connectivity

independent of N

g(rc) = g0exp[-2(rc-2b)/ξ]

Phys. Rev. B, 81, 155434 (2010)

The prediction of the percolation-threshold (critical-

resistor, hopping like) model and its relation to the

critical-percolation (phase-transition like) model

ℓn(N)

ℓn(σ)

Δσc

ℓn(σ) = -(aβ/ξ)Φ-β+ℓn(A2)

Φ N1/3 <β < 1

Phys. Rev. Lett. Comment, 106, 079701 (2011)

t = 4.54

How is it then that

in so many cases

we see a

percolation

behavior (even a

universal one)?

g(rc) ≈ g0exp[-2(Bc/N-) 1/3/ξ]

The origin of the staircase model

p

Log(σ)

pi,

Zi

g(rij) = g0exp[-2(rij)/ξ]

J. Phys. D: 42, 064003 (2009)

pc→ Bc/Z

The lattice and continuum

composite models

D

ξ

The effect of tunneling in lattices and the continuum

Phys. Rev. B, 82, 134201 (2010)

Staircase observed in N990/Polyethylene Composites

0.45 0.50 0.55 0.60

10-5

10-4

10-3

10-2

10-1

(

cm)-1

CB volume fraction

t=1.90 , xc=0.415

t=1.70 , xc=0.465

t=2.05 , xc=0.500

t=1.95 , xc=0.570

Conclusions• The observed σ(Φ) behavior is a series of percolation

transitions, the stairs.

• In the lattice these have to do with the series of lattice neighbors. In the continuum these have to do with shells of neighbors that result from an RDF with peaks.

• In the continuum the σ(Φ) behavior of each stair is that of a universal behavior while the envelope of the stairs yields a non universal percolation behavior which simply presents hopping.

• The experimentally observed Φc is either of a “universal” stair or a result of the fact that the data do not go to Φ = 0. Percolation is well confirmed if both, a universal behavior and a non hopping behavior are exhibited by the data.

• One can get a simple universal percolation behavior if all the resistors are nearly equal (non divergent distribution) and all belong to a random network (as for anisotropic particles).

The End

RF transmission (“AC Conductivity”) at 80 MHz

0.45 0.50 0.55 0.6010

-5

10-4

10-3

10-2

10-1

100

(

cm

)-1

CB volume fraction

t=0.4 , xc=0.415

t=0.7 , xc=0.465

t=1.65 , xc=0.500

t=0.45 , xc=0.570

f(g

)

g

Area = p = 2pcArea = pc

What if we have a distribution of (or r0 = 1/g) values?

RL <r>(p-pc)-tun

<r> = gc∫1(1/g)f(g)dg.

p[gc∫1f(g)dg] = pc

f(g) = (1-)g- yields that:gc = [(p-pc)/p]1/(1-)

<r> (p-pc)-/(1-) (p-pc)

-(t-tun)

Universal and Hopping Interpretations in

XC72-Polyethylene Composite

1.6 1.7 1.8 1.9 2.0 2.1 2.2

10-4

10-3

10-2

10-1

100

(

cm

)-1

(CB volume fraction)-1/3

, x-1/3

(b)

0.01 0.110

-4

10-3

10-2

10-1

100

(

cm

)-1

CB , x-xc

measured data

t=1.95 , Xc=0.1

Non-Hopping but Percolation Behavior

0.24 0.26 0.28 0.30 0.32 0.34

10-10

10-9

10-8

10-7

10-6

(Si content)-1 (vol. %)

-1

(Si content)-1/3 (vol. %)

-1/3

0.015 0.020 0.025 0.030 0.035 0.040 0.045

Co

nd

uc

tiv

ity

(

cm

)-1

Phys. Rev. B, 83, 035318 (2011)

t

cpp )(

Percolation Transition

t = tun = (D-2)+

(1-p)

tun (D = 2) ≈ 1.3

tun (D = 3) ≈ 2

The LNB Model of the backbone

R (≥ r0L1)RL= R (L/)/(L/)D-1

RL (p-pc)-[ +(D-2))] (p-pc)

-tun L1(p-pc)/pc =1 L1 (p-pc)-1

(p-pc)-

ζ ≥ 1

r0L1L

S&A 1992

The electrical conductance in composites

and the analysis of the conductance network

One expects intuitively and we can show rigorously that

V-Vc = (N-Nc)(vHC) (vol.%) in the continuum, is equivalent to ps-psc in lattices

TUNNELING (Wiesendanger, 1994)

LL

L ≤ 10 nm

σ(r) = σ0exp[-2r/L]

E

Resistors Equivalent Network

(Ii) = (gij)(vi-vj)

R = (v1-vN)/I

I1 = -IN = I, Ii (i # 1,N) = 0

Gii = Σgi,j, Gij = -gji

(I) = G (v)

(v) = G-1(I)

Phys. Rev. B, 28, 3799 (1983): 162 Q

The critical behavior of the resistance of

pencil generated line-segments system

1/gl = A0l

l

J. Water Resources Research 29, 775 (1993) : 101 Q

Non Universal and Hopping Interpretations

0.45 0.50 0.55 0.6010

-6

10-5

10-4

10-3

10-2

10-1

100

(

cm)-1

CB (volume fraction)

t=10.95 , xc=0.305

(a)

1.14 1.16 1.18 1.20 1.22 1.24 1.26 1.28 1.30 1.32 1.34

10-5

10-4

10-3

10-2

10-1

(

cm

)-1


, x-1/3

(b)

The model of Kogut and Straley

f(g) = (1-)g-

• <r> =gc∫1 [f(g)/g]dg |(gc

- -1)|, gc = [(p-pc)/p]1/(1-)

• we have here the three types of <r> = F(gc)

behaviors; strongly-diverges for 0 < < 1, logarithmically diverges for = 0 and it will not diverge for < 0.

• Correspondingly: <r> (p-pc)-/(1-),

• <r> |log(p-pc)| and <r> = [(1-)/(-)] = const.

The basics of non-universal behavior

RL <r>(p-pc)-tun

<r> = gc∫1(1/g)f(g)dg.

p[gc∫1f(g)dg] = pc

Example (K&S, 1979): f(g) = (1-)g- yields that:

gc = [(p-pc)/p]1/(1-)

<r> (p-pc)-/(1-) (p-pc)

-(t-tun)

(<r> = [(1-)/(-)])

“RDF” of the polymer shells

Evidence for stairs in MWCNT-polymer composites

Some idea about the materials

Evidence for stairs in MWCNT-polymer composites

Conductivity and viscosity

RDF and Cumulative coordination number

RDF in the Continuum

Statistics of the t values in CNT-polymer

composites

Two presentations of the same data

Percolation interpretation “hopping” interpretation


0.4 0.5 0.610

-6

10-5

10-4

10-3

10-2

10-1

100

Co

nd

uc

tiv

ity

(c

m)-1

x (volume fraction)

1.15 1.20 1.25 1.30 1.3510

-6

10-5

10-4

10-3

10-2

10-1

100

101

Co

nd

uc

tiv

ity

(c

m)-1

x-1/3

(x volume fraction)

-1.6 -1.4 -1.2 -1.0 -0.8 -0.610

-6

10-5

10-4

10-3

10-2

10-1

100

Co

nd

uc

tiv

ity

(c

m)-1

log10

(x - xC)

The physical and mathematical resemblance of the

percolation and (CPA) hopping behaviors:

In the classical hopping (CPA) model :log σ = -δc/ξ+log A1 = -(aβ/ξ)x-β +log A1 1/3 ≤ β ≤ 1

In our percolation model: log σ = t log(x-xc) +log A2

and we saw (from the Excluded volume argument) that

t = tun+d/ξ -1 so that for t >> tun we have that t ≈ -(x-β)/ξwhere 1/3 ≤ β ≤ 1 and log (x-xc) < 0.

• The two models yield then quite a similar dependence on x. Experimentally; since usually x >> xc these two are usually non distinguishable. Can we still distinguish between them?

Presentation of the CPA results in the case of spheres

CB-polymer composites

•Non universal value of t ≈ 6 for low structured CB composites

(p-pc)t

t = 6.4

vc = 39 vol.%

p = v/vc

R (≥ r0L1)

Phil. Mag. B, 56, 991 (1987): 142Q

RL= R(L/)/(L/)D-1

Non Spherical Fillers

Graphene composites

effects of particle shape asymmetry: spheroids (ellipsoids of revolution)

GTN for colloidal conductor-insulator composites

prolate spheroid (a/b >1)

oblate spheroid (a/b <1)

σ(v) behaviors

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.910

-7

10-6

10-5

10-4

10-3

10-2

v content

2

1

Co

nd

uc

tiv

ity

(

cm

)-1

Ge film

Ni-SiO2

Typical behavior of all composites

Percolation Threshold

Non Universal behavior

Geometry of Local Resistors

RV

IRV-like

(microemulsions)

PT

Local

connectivity

criterion?

IRV

(porous media,

impurities in SC

model)

A percolation threshold (critical resistor)

system vs. a critical percolation (phase

transition like) system

ℓn(N)

ℓn(σ)

Δσc

log σ = -(aβ/ξ)x-β+log A2

Critical Path AnalysisOne labels the conductances (the interparticle

distances) in the system in descending order

until one gets percolation i.e., one finds the

thinnest soft shell δ/2 that still provides

percolation for a given Φ.

Actually, we implant hard core particles and

then attach to each a shell of thickness δ/2.

We check then how many particles, or how

high a Φ is needed in order to get a

percolation path.

The agreement of the CPA with simulation of

the resistors network suggests that one finds

the largest σ(Φ) without having to calculate

the entire resistor networks.

This works because of

the very wide resistor

distribution values

The tunneling percolation staircase

p

Log(σ)

pi,

Zi

tun= 1.9

J. Mod. Phys. B, 18, 2091 (2004): 52 Q


p

Log(σ)

pi,

Zi

tun= 1.9

Int. J. Mod. Phys. B, 18, 2091 (2004): 52 Q

Two interpretations of the full set of data points

LATTICE PERCOLATION

ps

pb

S

pc

r0

J. Water Resources Research 29, 775 (1993) : 101 Q

A typical continuum system

No p here

Local

connectivity

by partial

overlap

The End

The Staircase in the AC measurements

0.45 0.50 0.55 0.60

108

109

1010

Y A

xis

Title

X Axis Title

C

t=0.4

t=0.7

t=1.65

t=0.45

Fluctuations in “experimental “ data

Two presentations of the same data

The volume and excluded volume

of the capped cylinders

v = (4/3)(W/2)3+(W/2)2L

<Vex> = (32/3)(W/2)3+8π(W/2)2L+4(W/2)L2<sinθ>

<sinθ> = /4

Nc = (Bc/v)v/<Vex> (Bc/v)(W/L)

when L<<W (0 ≤ θ ≤ )

Bc found to vary relatively

little with aspect ratio and is

practically independent of

anisotropy

The covered volume will be

proportional to Ncv W/L

Phys. Rev. Lett. 52, 1465 (1984): 186 Q

Electro-Thermal Switching: the „global‟ effect

• an increase of the macroscopic resistance above a threshold bias

• a stationary “switched” state is reached within 30 sec

-20 -10 0 10 20-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

Cu

rre

nt

(A)

Voltage (V)

t~1 sec

t=30 and t=300 sec

Phys. Rev. Lett. 90, 236601 (2003): 20 Q.

V=3

V

V=6 V V=8 V

Making the Audio Disc

HISTORY

• 1982 RCA Laboratories, Princeton, USA

• Request: High conductivity, High elasticity.

• Question: Carbon Black supplier?

• A) Cabot Corporation, Boston USA

• B) Akzo Chemicals, Amersfoort, The Netherlands

Low structure or high structure ?

Video Disc Radius (In)

The “lost” signal (Percolation Threshold and Anisotropy)

Phil Mag. B, 56, 991 (1987): 142 Q.

Overcoming the “Anisotropy”

The Question is:

How to express

conceptually and

mathematically the

feeling that

elongated, not

parallel, particles will

yield a higher

conductivity (i.e.

connectivity) for the

same CB content?

Cylinder like objects

Phys. Rev. B, 28, 3799 (1983): 156 Q

Behavior of three types of composites that are

based on elongated carbon particles

Typical behavior of all composites

Percolation Threshold

the percolation picture implies the presence of a sharp cut-off, but tunneling does not

percolation versus tunneling

g(r12) = 0 for r12 > D+d

percolation

g(r12) ≠ 0 for r12 < D +d

tunneling

g(r12) = g0 exp[-2(r12-D)/]

tunneling does not

imply any sharp cut-

off

Carbon, 40, 139 (2002): 82 Q.

The concept of “excluded volume”

R

R

2R

Vex = v2D

The Excluded

Volume is the

volume in which

the center of the

“other object”

has to be in order

for the two

objects to, at least

partially, overlap

vNc = (v/Vex)(VexNc) = (v/Vex)Bc

We generalized v2D to other objects calling it “Excluded Volume”

Note: 2D = smallest Vex/v

for permeable objectsPhys. Rev. B, 30, 3933 (1984): 290 Q

NVex = B: the average

number of overlapping

objects, i.e. the average

number of bonds per

object. Bc is expected

then to be a topological

invariant.

History of percolation Thresholds in

the Continuum (Sastry, PRE 2007)

Table 2. Key percolation studies in three dimensional random fibrous materials

Authors Year Reference Arrangement Objects Contribution Approach

Balberg, Binenbaum and

Wagner

1984 27 isotropic capped cylinders

First study of percolation of random objects in three

dimensions

Monte Carlo

simulations

Balberg, Anderson, Alexander et al.


Presented the conjecture that percolation threshold for a system of identical objects in three dimensions is inversely

proportional to excluded volume of one object

Monte Carlo simulations

Bug, Safran and

Webman 1985 28 isotropic

capped

cylinders

Confirmed excluded volume

rule and showed constant of proportionality equals one in slender rod limit

cluster

expansion

Neda, Florian and Brechet


Performed Monte Carlo simulations of 3D stick systems using correct isotropic

distribution and confirmed excluded volume theory


Yi and Sastry 2004 30 isotropic ellipsoids

Presented analytical

approximation for the percolation threshold of two- and three- dimensional arrays of overlapping ellipsoids

series

expansion

Yi, Wang, and Sastry 2004 31

Uniform distribution in x-y plane. Range of

distributions from parallel to random in z-direction

ellipsoids

Provided comparisons of cluster sizes, densities, and percolation points for two- and three-dimensional systems of

overlapping ellipsoids to investigate the range of applicability of a 2D model for predicting percolation in thin

three-dimensional systems


The generalization to the hard core caseIn considering

tunneling, the

tunneling range is

R-r = δ

The important

observation is that :

Фc = Ncv (Bc/ΔVex)v

r2L/[L2(R-r)]

[r/(R-r)](r/L)

If δ = (R-r) we

have that

Фc (r/δ)(r/L)

For L >> R

Vex ≈ 4L2R<sin Θ>

ΔVex ≈ 4L2δ<sin Θ>

For Oblate Particles

Asymptotically for a >> b, d and isotropic distribution:

ΔVex [(a+d)3 –a3] ≈ (a2d), v = πa2b, Фc b/d = (a/d)(b/a)

From the simulations:

Фc v/Vex ≈ (a/d)(b/a)1/2

Phys. Rev. B, 81,155434 (2010).

For discs of

radius a and

thickness b

Vex a3


p

Log(σ)

pi,

Zi

tun= 1.9

f(g

)

g

The RDF

The effect of a tunneling like distribution of conductances

t = 6.4f(g) = (1-)g-

<r> (p-pc)-/(1-)

f(g) g(ξ/d-1)

t = tun+(d/ξ)-1

t

cpp )( V V

f(g

)

g

A percolation threshold (critical resistor)

system vs. a critical percolation (phase

transition like) system

ℓn(N)

ℓn(σ)

Δσc

log σ = -(aβ/ξ)x-β+log A2

Critical Path AnalysisOne labels the conductances in the system in

descending order until one gets percolation

i.e., one finds the largest soft shell δ that is

needed to get percolation for a given Φ.

Actually, one implants hard core particles and

then attaches with a shell δ/2. One checks

then how many particles, or how high a Φ is

needed in order to get percolation.

The agreement of the CPA with simulation of

the resistors network suggests that one finds

the largest σ(Φ) without having to calculate

resistor networks.This works

because of the

very wide

distribution of

resistor values

Presentation of the CPA results in terms of percolation

0.0 0.1 0.2 0.3 0.40

10

20

30

40

50

60

Threshold Content C

Th

e e

xp

on

en

t

t

application of CPA formulas to real conductor-polymer nanocomposites


from the knowledge of vs f and from the filler geometry [a/b and

D=2max(a,b)] one can estimate the value of the tunneling factor

The physical and mathematical resemblance of the

percolation and (CPA) hopping behaviors:

In the classical hopping (CPA) model :log σ = -d/ξ = -(aβ/ξ)x-β +log A1 1/3 ≤ β ≤ 1

In our percolation model: log σ = t log(x-xc) +log A2

we saw (from the Excluded volume argument) that

t = tun+d/ξ -1 so that for t >> tun we have that t ≈ -x-β/ξwhere 1/3 ≤ β ≤ 1 and log (x-xc) < 0.

• The two dependence yield then quite a similar dependence on x. Experimentally; since usually x >> xc these two are usually non distinguishable. Can we still distinguish between them?

application of CPA formulas to real conductor-polymer nanocomposites


even if a/b changes over 6 orders of magnitude, the tunneling factor is found to be within the

expected range 0.1 nm < < 10 nm [G. Ambrosetti et al. Phys. Rev. B 81, 155434 (2010)]

Analysis of a resistors network

Critical Path AnalysisOne labels the conductances in the system in

descending order until one gets percolation.

One takes then the smallest conductance thus found

and determines its δ(Φ) value. The agreement

with simulation suggests that:

One can simply arrange the δ(Φ) values and

find directly the largest needed δ(Φ)

without having to calculate resistor

networks.

This works because of the very wide

distribution of resistor values

The excluded volume approach• The parameter that is known experimentally is the critical volume % of the “filler” Φc, so

the prediction to be made is how does Φc depend on the microscopic parameters.

• For spheres these parameters are D and so:

• v= (4π/3)(D/2)3, Φ = Nv, Φc = Ncv,

• Bc = (4π/3)(rc3-D3)N and δc = rc-D

• Case 1: rc >> D → Bc = (4π/3)(δc3)N = (4π/3)(δc

3)Φ/v

• → δc (Φ/v)-1/3 →

• Log ρ = α1 + β1Φ-1/3.

• Case 2: rc ≈ D → Bc ≈ (4π/3)(3δcD2)N = (4π/3)(3δcD

3)N/D = (3δc8)Φ/D

• Φ/v → Bc = 24 δcΦ/D → δc = BcD/24Φ

• δc DΦ-1

• Log ρ = α1 + β1Φ-1.

The predicted CPA conductivity

dependence on Φ

• For capped cylinders:

• v ≈ πR2L, Φ = Nv = πR2LN

• ΔVex ≈ πL2δc<sinθ >, NΔVex = Bc

• δc ≈ BcR[(R/L)<sinθ >]Φ-1 Φ-1

• In general then, δc Φ-β where in 3D: 1/3 ≤ β ≤ 1

CPA for spheres

The RDF

The End

VISCOSITY vs. CONDUCTIVITY

The differences between the different

types of elongated Carbon particles

For high structure

carbon black the

aspect ratio is of the

order of 10

The End

Recent results of the RF conductivity and

permitivity of Carbon Black composites

-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.47.5

8.0

8.5

9.0

9.5

10.0

10.5

11.0

11.5

measured 0.1GHz

measured 1GHz

measured 10GHz

calculated 0.1GHz

calculated 1GHz

calculated 10GHz

log

10(

(s

ec

-1))

log10

((p-pc)/p

c)

-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.40.8

1.2

1.6

2.0

2.4

2.8

3.2

measured 0.1GHz

measured 1GHz

measured 10GHz

calculated 0.1GHz

calculated 1GHz

calculated 10GHz

log

10(

)

log10

((p-pc)/p

c)

Conductivity exponent

0 2 4 6 8 10 12 14 16 18

1.0

1.5

2.0

2.5

3.0

3.5

t

f (GHz)

N990

Raven

XC72

Permitivity exponent

0 2 4 6 8 10 12 14 16 180.0

0.4

0.8

1.2

1.6

2.0

N990

Raven

XC72

s

f (GHz)

f(g

)

g

An illustration of the basic answer

Occupied sites

Empty sites

If resistors R3 >> R2 >> R1 are attached, the resultant resistance of

the system will be dominated by R2, i.e., for a given p, an R2 resistor

will yield a threshold resistance of the system (R1→ 0 and R3 → ∞).

Approach: Keeping p and going from Z1 to Z2 in the square lattice

Bc = zpc =pzc

The staircase in a lattice

p

Log(σ)

pi, Zi

The change of the average local

resistance upon threshold approach

(the bypassing network)

• 1) Start from p > pc

• 2) Consider the g‟s in descending order

• 3) The largest pc of them yield percolation

• 4) The smallest g in this subset is gc.

• 5) If there is a distribution, gc decreases as p → pc.

• 6) At p = pc the resistor of smallest g must be included

• 7) If it is g → 0 it may result that <r> → ∞

UPON p→pc NOT ONLY THE NETWORK SHRINKS BUT THE AVERAGE RESISTANCE INCREASES AND MAY DIVERGE

A typical (pencil generated)

continuum system

No p

no f.

and N is not

informative

Phys. Rev. B, 28, 3799 (1983): 156 Q

The lattice composite model

Statistics and conductance of a composite system

RDF of spherical particles

Staircase structure in nano Graphite composites


0.45 0.50 0.55 0.6010

-6

10-5

10-4

10-3

10-2

10-1

100

(

cm)-1

CB (volume fraction)

t=10.95 , xc=0.305

(a)

1.14 1.16 1.18 1.20 1.22 1.24 1.26 1.28 1.30 1.32 1.34

10-5

10-4

10-3

10-2

10-1

(

cm

)-1


, x-1/3

(b)

Ni-SiO2

10-4

10-3

10-2

10-1

100

0.20 0.30 0.40 0.50

10-4

10-3

10-2

10-1

100

Co

nd

uc

tiv

ity

(c

m)-1

x (volume fraction)

t = 3.3

Two fundamental questions

1) What determines the

percolation threshold?

2) What determines the

shape of the σ(x)

dependence?

Another a priori unexpected behaviors

0.4 0.5 0.6 0.7 0.8 0.910

-1

100

101

102

vol. fraction (v)

2

1

Co

nd

uc

tiv

ity

(

cm

)-1

fractional Ni content

Percolation and tunneling -percolation

Various carbon blacks

Two Conclusions The understanding of the conductivity dependence

on the content of the carbon filler in terms of

percolation theory is based on:

σ (Φ-Φc)t

• 1) The larger the aspect ratio and the isotropy

the smaller the percolation threshold (the

smaller the content of carbon particles needed

for the onset of large conductivity), the

“excluded volume” argument.

• 2) The Behavior of the conductivity is

determined by the distribution of the tunneling

conductances (or interparticle distances) which

is manifested by the deviation of the critical

exponent t from its universal value.

The origin of the staircase model

p

Log(σ)

pi,

Zi

g(rij) = g0exp[-2(rij)/ξ]

J. Phys. D: 42, 064003 (2009)

pc→ Bc/Z

The NLB Model

A model for the links

R (≥ r0L1)RL= R(L/)/(L/)D-1

(p-pc)-[ +(D-2))] (p-pc)

-tun(L1 (p-pc)

-1)

An illustration of a Graphene composite

v = πR2d

Vex = π2R3

v/Vex d/R

The Graphene-composite problem

• For example, for high density spheres

• δc = BcD/24,

• for cylinders δc ≈ BcR[(R/L)<sinθ >]Φ-1 and

• for discs Bc = π2(3b2δ+3bδ2+δ3)N so that

• δ/t ≈ Bc/(3πФ), or ≈ Ф ≈ [t/5δ)], or . δ ≈ tBc/(3πФ),

The Tunneling Percolation Staircase€¦ · The Tunneling Percolation Staircase Isaac Balberg Racah...

Documents

Transcript of The Tunneling Percolation Staircase€¦ · The Tunneling Percolation Staircase Isaac Balberg Racah...